|
| The Digital Hive |
| 23-year-old Irishman Ian Clarke, as a student at the University of Edinburgh in Scotland, developed Freenet, an anonymous P2P file sharing network. |
| Writer: Spiff |
Jan 06 95 |
|
|
|
Ian Clarke isn't a household name yet -- emphasis on the yet.
Clarke, if you're not the sort that closely follows technology news or developments in the realm of file-sharing software (and let's face it, not many people are that sort at all), is the 23-year-old Irishman who, as a student at the University of Edinburgh in Scotland, developed Freenet, a piece of technology that, if it becomes as virally popular as the Napster and Gnutella software applications, will revolutionize not only the Net but also the very concept of intellectual property.
That's a bold statement, but it's not an exaggeration. Why the commotion? Because Freenet works like Gnutella on steroids. File-swapping technology has created an uproar because it allows users to easily trade copyright-protected information -- MP3s being the format most commonly discussed (and, arguably, traded). As broadband access increases and as the world's telecommunications infrastructure grows stable and large enough to sustain massive transfers of data, it's feasible that movies and bootlegs of television programs may become as widely traded as MP3s are now. Imagine being able to access any piece of music, literature or cinema ever made at the stroke of a key. Imagine having access to any piece of information, from government documents to naked pictures of Jenny McCarthy, with minimal effort and minimal wait. Now imagine the creators of that information not being compensated for the use of any of that data. That's what file-sharing technology enables, albeit in an infant stage right now.
The major representatives of intellectual property owners (the Recording Industry Association of America, the Motion Picture Association of America, and the National Music Publishers Association) are currently engaged in a vicious battle to stop the spread of what they consider to be digital piracy. Heretofore, the one method of combating unauthorized intellectual property distribution has been to identify a pirate through his or her Internet Protocol (IP) address, a series of numbers that communicates the location of a person's computer on the Internet, and to shut them down, either by legal injunction or law suit. Set up an FTP server for people to grab MPEG files, and in short order, a member of the RIAA or MPAA can have you shut down and, possibly, arrested. If you're a corporation whose software enables the unlicensed use of intellectual property, you could even find yourself being sued. Just ask the folks over at Napster and MP3.com.
Freenet circumvents this threat by encrypting the identities of its users, making all participants in the system completely anonymous. Further, unlike Napster, it is decentralized, meaning that there's no corporation channeling the data through one set of servers, thereby making themselves potential targets of litigation. In other words, Freenet is a freedom-of-information advocate's dream come true and -- in the eyes of the RIAA and MPAA -- the entertainment industry's worst nightmare.
It's no surprise, then, that Clarke has found himself under the media microscope lately. He has been the subject of numerous articles and has spent a lot of time enduring questions from members of the music industry (most notably at the MP3 Summit in La Jolla, Calif., and at a Digital Coast dinner / conference). He's also being followed around by a TV crew from the CBS news program 60 Minutes II; the show plans to air a segment on Freenet this fall. Thankfully, Clarke's not so inundated with press requests that he couldn't take time to speak to the Webcasting community. He spoke with RadioSpy's Sean Flinn via phone from the United Kingdom on, fittingly enough, July 4.
Sean Flinn: The first questions I have are directed at people who really are unfamiliar with Freenet and unfamiliar with who you are. Can you take a minute to introduce Freenet and introduce yourself? Tell me a little bit about your background and what you do for a living.
Ian Clarke: Freenet is a protocol, a system that allows people to distribute information on the Internet without fear of censorship. And it does that by providing complete anonymity to both those people placing information onto the system and to people who are [downloading] information that is stored on the system. It's also impossible to identify where a piece of information is stored on the system -- in fact, a piece of information can be stored in more than one place, which makes it very difficult -- if not impossible -- to remove information from the system.
In terms of me, I developed Freenet as part of my fourth-year project while studying artificial intelligence and computer science at the University of Edinburgh, in Scotland. I completed that design in July of 1999 and released the design onto the Internet along with an invitation to people to help me make this design a reality. And really, it's gone from there. We released our first prerelease version in March [2000]; we released another version several weeks ago; and we're coming up to what will be the third version of Freenet.
This was a school project? What grade did you get on it?
I got a B.
Wow. What would it have taken for you to get an A?
I think it would have to have been less controversial.
Gotcha. So people are already a part of the system, and it's actually working?
It is. Yes indeed. It's been working since mid-March, although it is still very much in development. These are prereleases. We don't claim that Freenet is complete -- although it does work, and it is useful, and it does distribute information anonymously, and it does so quite efficiently. But there're still a lot of features that we want to implement before we're willing to describe it as "complete."
What was your inspiration for developing Freenet? I mean, I know that you were studying artificial intelligence at U of E, but what specifically led you to develop this sort of technology?
I think that there were probably two things. I mean, there's an ideological reason for it, and there's a kind of technological reason for it. In terms of the ideological reason, I believe in total freedom of information. And I think that if you do believe in total freedom of information, you can't have half measures. You can't say, "This is permissible, and this isn't." Now, I realized that it would be quite difficult to convince other people of this, and so the situation was unlikely to come about through democratic means. But I also realized that I could write a piece of software that would effectively force the issue.
My motivation from the technical side was, firstly, really, I was fascinated by the idea of complex systems, which are formed from simple individual entities all cooperating. An example would be an ant's nest, whereby all of these ants are following relatively simple rules, yet they all work together to make this effectively a kind of meta-organism, which is the ant's nest, which can feed itself and reproduce and defend itself. So I was fascinated by that idea, and I was very interested in trying to apply that to a computer system. And by combining these two ideas, I essentially came up with Freenet.
How many ants are in the Freenet anthill by this point?
It's just a suspicion because there's no way to take a census of people running Freenet nodes -- that's quite deliberate -- but we've had about 100,000 downloads now, and we have about 1,500 downloads a day. There's no guarantee that every single person who downloads the software is actually running a node, but it certainly gives an indication.
You mentioned there were some other features that you wanted to add to Freenet. What kinds of features?
At the moment, you can look at Freenet as [being] a little bit like a file system. And just like a file system, in order to retrieve a file, you need to know the file name. Now, unlike a file system, there's no way to get a list of all the files, and there's no way you can search through the files on the system. So you need to now the exact name of the file you're looking for. We want to address this issue. We're working, at the moment, on a system that will allow you to search Freenet, to do a fuzzy search of the system for files. We also want to install a mechanism whereby people can update information on Freenet. At the moment, once you place information on Freenet, it stays there, and you can't update it. So we also want to create a system whereby people can update information. I would say those would be the two core improvements that I hope we'll see over the next couple of months.
When do you expect Freenet to exit its prerelease stage and be offered up fully for public participation?
Well, I don't think there will ever be a grand fireworks kind of [release]. The paradigm with open-source development tends to be extremely conservative in terms of setting deadlines, to be extremely conservative in terms of describing your software as "finished." If you look at Linux, there are pieces of software on Linux that have been around for a long time and which are extremely stable, but the authors still refuse to say that they're finished. So I don't think that we'll have any grand announcement that Freenet is finished. I think it will be a gradual process of refinement and improvement.
Does someone basically become a part of Freenet by downloading the software from the Freenet home page and running it?
Yes.
And is it fairly intuitive? I mean, can the layman pretty easily pick up the basics of it and run it?
No. Not right now. We have, in a sense, made it deliberately complicated to install, primarily because, as I said, it is still in development. People who are more computer literate tend to give more valuable feedback, so we wanted to raise the barriers to entry in the short term, during the debugging and development process. But in the very near future, I think we will probably make it very easy to install. In principle, there are no reasons why it can't be as easy -- if not easier -- to install than Napster.
And the technology of Freenet, like Napster and Gnutella, enables direct file swapping between users, correct?
Not exactly. If you look at Napster and Gnutella, the way they work is that you have files on your computer, and you're sharing them with the world. Freenet doesn't quite work like that. With Freenet, you place files into Freenet, and Freenet decides where they should go. And similarly, you don't know what files Freenet is placing on your machine.
I guess what I was aiming at was -- well, are you familiar with technology like SHOUTcast and Icecast that have made it possible for people to run Internet radio stations?
Yes, I am. Yes.
Is there any way for people to use Freenet's technology -- especially the anonymity it affords its users -- to transmit streaming audio? The benefits of that would really seems to pay off for people in political systems like China's, where the uncontrolled freedom of information could have dramatic political consequences.
Well, at the moment, you can play audio on the system. You can place an MP3 on the system. But it's not like a live transmission. It's not radio. It is interesting that you bring this up because at the moment, I've been approached by a gentleman who's setting up a company -- he's asked me not to go into details about what he's doing -- but he's actually going to work with us to implement essentially this exact functionality in Freenet. The ability to use Freenet to set up a radio station on which you could do live broadcasts, but in such a way that would still be very efficient -- much more so than technologies like SHOUTcast. It may not be quite as secure, in terms of anonymity, as normal Freenet, but it would offer a level of anonymity.
But the main benefit of anything -- and this is kind of something that people frequently don't pick up on -- is that Freenet is actually a very efficient distribution mechanism. Are you familiar with Akamai? Well, Freenet does what Akamai does but, basically, takes it several steps further. Let's say a piece of information is very popular in the U.K. Freenet will actually move that information to the U.K. so that people in the U.K. can access it faster. Similarly, if a piece of information suddenly becomes extremely popular, Freenet will duplicate it to spread the load, so you don't get a kind of slashdot effect, whereby a server goes down if it contains a piece of information that suddenly becomes popular. So, in many ways, Freenet actually addresses a lot of efficiency issues with the Internet and with technologies like SHOUTcast and RealAudio in addition to offering a censorship-free forum. And streaming media over Freenet will benefit from those efficiency improvements.
That'll be huge. Like I said, the political implications of that are astounding, just because of the anonymity it affords to people speaking out against their governments in systems that are very hostile to dissent.
Well, that's right. Freenet now would support anonymous speech that could not be removed by a government. But yeah, in terms of live broadcasts, it is an interesting area.
But still down the pike a little bit?
It is down the pike a little bit, but like I said, this private company, which is well funded, is interested in working with the Freenet developers to develop this. So it will probably happen quite quickly, given that it's got people really behind it and really pushing it.
Let's talk a little bit about the Freenet ratings system. I know Freenet is set up to distribute and redistribute popular content to places where the demand for it is great, to make it easier to access. But it also sets up a system whereby unpopular content is ejected. Can you qualify what is meant by "unpopular"? Is it based on a ratings system where, for example, people can listen to MP3 files and report, "I like this song," or, "I hate this song," or, "This MP3 is not what it purports to be. It's a 'Napster bomb'"?
It's very simple at the moment. If you listen to or if you access a file or read a file from Freenet, you're effectively voting for it. So popularity is determined purely by how many people access it. Now, people occasionally suggest having a ratings system whereby people can say, "This is good," or, "This is bad." But we don't want to do that because that would lead to censorship by majority, and that's censorship just the same as any other kind of censorship. We don't want that kind of situation. But what we will allow -- this is another one of the features -- is that if you download a piece of information and it's not what it purports to be, you can undo your vote, so to speak. So you can't actually vote negatively for a piece of information. You can't say, "This is bad." But what you can do is undo a positive vote. So you can't do any damage.
I guess the concern that I've heard mentioned by some of my colleagues in the online music space is that a rare recording by an underground band might not be accessed very heavily and, therefore, wouldn't be as popular as some other content and would be ejected from the system.
Well, that is certainly potentially true, but you really have to see content being ejected from Freenet as a last resort. Freenet ejects content if it is taking the place of something else that is more popular. So it really is a last resort as opposed to something that happens all the time. So something would have to be really, really, really unpopular -- like have about one person access it a year -- for it to drop of the system. I can't state that categorically, but you know, something would have to be very unpopular for it to drop off Freenet.
I just know that people are concerned that it will be nothing but a bunch of Britney Spears MP3s, and they won't be able to get anything that they're really looking for.
No, you're talking about Napster there. No, Freenet should be quite good. Again, some of these things are "Let's wait and see." We're kind of looking at Freenet with almost as much interest as other people because, up until now, it's all been theoretical. We have developed a simulation very recently -- in fact, it was just completed a couple of days ago by two of our developers -- which will allow us to really pick Freenet apart and tweak things here and there, so that should lead to a lot of improvements, now that we have that tool. But yeah, it's not going to be a case of it just being Britney Spears. Really, something needs to be very unpopular for it to drop off Freenet. And even if it does, somebody could just put it back on the system again, anyway.
People also say, "Wouldn't it be great if Freenet were a permanent repository of information?" And we're like, "No, it wouldn't be great because it might fill up, and people wouldn't be able to add stuff." But our counterargument is that Freenet doesn't purport to be that. If you have information, and let's say you put it on Freenet, and nobody accesses it, and it drops off -- you can just put it on again. It's up to the person placing the information on Freenet. If they want to make sure it stays there and if it's not a very popular piece of information, then all they have to do is periodically put it back on. And they can do that while retaining their anonymity, etc.
I've seen something in the Freenet FAQ suggesting that the anonymity protection and the encryption aren't perfected yet.
Yes. This is another example of the kind of conservatism that is prevalent with computer scientists, in that there are very different standards of anonymity. Now, one standard of anonymity is that it is mathematically impossible to identify people. It is not mathematically impossible to identify people on Freenet. What we would say -- and what we hope and what nobody has yet been able to dispute -- is that it is, for all practical purposes, impossible to identify somebody on Freenet. Even if you're the American government. Even if you're the Chinese government. It is, for all practical purposes, impossible. But it is not a mathematical impossibility, and I think that's probably what you picked up on. It's just ultra-caution on our part. We don't want to upset these people who are anonymity freaks, who are just stunningly paranoid. But in layman's terms, Freenet is practically 100-percent anonymous.
You've been under quite a bit of media scrutiny recently and under attack from members of the commercial media -- especially record label executives who've said a lot of outrageous things about what you've done with Freenet. My interpretation of all of this is that it's made you the face of Freenet, and it's focused attention all on one source. Do you foresee any attempts at legal intervention on the part of intellectual property owners and antipiracy activists aimed at you, using you as a centralized point for attack?
I've actually been surprised by the lack [of these efforts]. I mean, I think Napster has taken a lot of flak, but even Lars Ulrich [of Metallica, the most vocal artist to date to come out against Napster and Internet music piracy], who is, in the eyes of some, my arch enemy, is quite moderate on the one or two occasions I've seen him mention Freenet in interviews. So, maybe you've seen things that I haven't, but I've been surprised by the lack of wild, outrageous statements. I've seen one or two directed at Napster, but I haven't seen much really directed at Freenet.
The most outrageous thing I've seen was in an article reporting on an interview you gave at a recent Digital Coast gathering, where somebody said that what you're doing with Freenet is akin to making robes or hoods for the Ku Klux Klan.
Unfortunately, [the reporter] didn't print my retort. I asked whether they thought that outlawing the manufacture of white hoods would solve the racism problem. Unfortunately, they didn't print that. OK, that was probably my most hostile reception, and that was very much the exception as opposed to the rule. I wasn't really perturbed by that very much. I almost just sat [quietly] up on stage because some people in the audience -- all of whom, to the best of my knowledge, were music industry people -- would say one thing, and when I would go to give an answer, these other people would shout back at them! So you really saw both sides of it, even at that, and that would be the most extreme example.
I remember also, at the MP3 Summit in San Diego, one of your responses to questions and concerns about the theft of intellectual property was, "If your business model is selling water in the desert and it starts to rain, you'd better find a different business model." And we've had a lot of attention from record executives about Freenet, but what other industries -- business models -- do you see being impacted by anonymous, decentralized file sharing?
Well, it's only a matter of time before it hits the film industry. It's purely a question of download speed, and that's increasing all the time. You could probably figure it out with a pen and paper. If you graphed the kind of average download speed -- I mean, two years ago it was 56k modems. Now, a lot of people have cable -- and I'm not sure what speed cable is, but you could probably quite easily draw a graph that shows [how fast] the bandwidth into homes [is increasing], and you could correlate that graph against the bandwidth required for streaming television-quality video, and you could probably predict down to the month exactly when people could start distributing films in the same way that people are currently distributing MP3s. I should do that math. I'm not very good at math, though. Maybe you can do that math.
I'm not very good at math, either. That's why I'm a journalist. [Laughter] Let's get doomy and gloomy for a second. Are there any "nightmare scenarios" for Freenet? Can you see anything coming down the pike that could be a complete nightmare for you and the people involved with Freenet?
Yeah, well, if they tried to shut down the Internet, that would probably pose some serious problems for Freenet. But short of that, I think we can probably work our way around any sanctions.
One more plausible sanction they could take against Freenet would be banning encryption. Now, that would kill e-commerce, so it's unlikely to happen in many countries. But politicians can be very, very stupid, and for example, in the U.K., they're now bringing out a bill called the Regulation of Internet Powers bill, which is -- talk about shooting yourself in the foot; this is like taking a nuclear weapon to your foot -- they have effectively brought out this law whereby if you have an encryption key, the police can demand that key from you in order to decrypt information that you have. Not only that, it will be illegal for you to tell anyone that they've done that. That, in effect, makes encryption useless in the U.K., in some senses. And if they go ahead with it, e-commerce companies will be leaving the U.K. like rats jumping off a sinking ship.
That's kind of ironic because some countries are going in the opposite direction. Ireland, for example, has made it illegal for the government to even try to break encryption. France, which used to have quite draconian laws regarding anonymity and encryption, has now relaxed them totally. So the U.K. is really just a classic case of dumb-ass politicians.
Well, we've had our fair share of it over here, what with the laws against the export of strong encryption technology.
Well, yeah. But I think we've occasionally looked smugly over at you in terms of those laws, and unfortunately, this one is about 20 times worse. So it's kind of Americans' turn to look smugly at how stupid our politicians are.
It'll go back and forth, I'm sure.
They're competing! But yeah, that is plausible -- that they might ban encryption -- which would be kind of irritating. But the thing is that there are also techniques to disguise the fact that information is encrypted. There are ways, for example, that you can place encrypted information into an image which itself is unencrypted, and you wouldn't actually be able to tell that there's encrypted data stored in that image. So there are even techniques that can be used to get around really stupid laws like that. It is perhaps slightly sensationalistic to say that in order to completely shut down Freenet in such a way that there is nothing we could do to adapt it, you would need to shut down the Internet.
Which is pretty impossible, especially given what the original intent of the Internet was.
Part of the problem, though, is that although the original intentions were to make it decentralized and robust, a lot of the structures built on top of it since then have not been. In some ways, Freenet can be viewed as an attempt to return to the original intent of the Internet, which was this system that roots around censorship and all of these clichés that were never actually true until, perhaps, now. And the original Arpanet was [designed for the eventuality of] nuclear holocaust, when the only things that would survive would be cockroaches and the Internet. Cockroachnet. [Laughter.]
But since then, the Internet became very centralized, not very robust at all. In fact, the Internet occasionally goes down, in that the domain name system that the Internet relies on periodically has been known to break. And that's not even with a nuclear war. That's just normal usage. So the Internet, at the moment, doesn't really live up to the original claims of an indestructible system. I think Freenet is moving back in that direction.
Now let's get a little pie-in-the-sky for a moment: How would you describe Freenet functioning in an ideal manner? What's the most positive thing that you can see happening as a result of the proliferation of this technology?
In a sense, I think that ultimately there is the potential for Freenet to replace the World Wide Web. Now, at the moment, it doesn't have anywhere near that functionality. Nothing we've even proposed at the moment will give functionality akin to the World Wide Web in terms of dynamic CGI and stuff like that. But way down the line, it would be theoretically possible to have dynamic Web pages. I guess the ultimate thing in Freenet is for it to be as popular as the World Wide Web, used in the same way as the World Wide Web and maybe even replace the World Wide Web. But, as you said, that's complete pie-in-the-sky [thinking]. But if that happened, we would basically be living in a world of complete freedom of information, and I think that would be a great thing. Then people would have to learn how to decide what's good and what's bad and what information they should trust and what information they shouldn't trust, as opposed to the situation now, where everyone believes what Rupert Murdoch tells them.
I guess the concern then becomes -- and it's almost impossible to conceive of because it's somewhat antithetical to the core idea of Freenet -- how do people develop a business model to capitalize on complete freedom of information?
Well, it is funny you should say that, but that's exactly what I'm doing [with Upriser, a company Clarke mentioned during his keynote address at the 2000 MP3 Summit]. That's basically what Upriser is doing. We're developing a business model which will allow artists to profit from their creative output without relying on copyright -- in a sense, allowing artists to benefit from what's currently referred to as piracy and actually allow artists to do a lot better than they're doing at the moment. I mean, if you look at the music industry at the moment, practically the only people in the music industry who don't do very well are the consumers and the artists. The vast majority of artists don't make a penny from copyright. Even artists who are moderately successful, as Courtney Love mentioned [in a now infamous speech wherein she declared her intention to leave her record label and make her music available online -- which she did], even if you're a moderately successful band, you still don't make a penny. All of the money goes to the music industry. My hope is, with Upriser, that we will develop a business model that will benefit the artist, act as a democratizing force, so perhaps you have not as many Spice Girls but a couple more Radioheads. Which would be a --
A far better world.
[Laughter] It would be a better world, and there would be peace and harmony! But we'd all be quite depressed. So it's a lot to kind of claim that getting rid of copyright will lead to peace and harmony in the music industry and artists doing well. But I think that the business model we're developing -- and you can judge for yourself in about 40 days [as of the July 4 interview], when we release it -- will really do very, very good things for the artist. It's a revolutionary approach. So wait and see on that one. But yes, it is possible to build a business model around total freedom of information, and I'm planning to prove it.
My final question for you, then, has to do with the other network systems similar to Freenet: What do you think of Napster and Gnutella and Sealand?
We've spoken to Sealand. Some people on our mailing list were like, "Oh, wouldn't it be great to get a node on Sealand?" And I didn't say anything at the time, but they e-mailed us back and said, "Surely the point of Freenet is that you don't need Sealand in order to run a Freenet node." And I think they were dead right. In a sense, Freenet should make Sealand redundant, and it sounds like Sealand are quite content with that situation. They seem to be very positive about Freenet.
Napster, I guess, was first, but there are a lot of differences between Napster, Gnutella and Freenet. They're all kind of batched together, but their geneses were quite different. Shawn Fanning's motivation with Napster was to help people share music. My motivation was all about freedom of information and ants' nests. So really, their geneses were totally different. I'm not sure who started working on theirs first, but I kind of completed my design several months before I heard about Napster. So it wasn't really a case of me looking at Napster and saying, "Oh! I can do it better." I actually started on Freenet, conceivably, before Shawn Fanning started on Napster. The obvious difference is that Napster is centralized; Napster doesn't afford anonymity, basically doesn't do any of the things that make Freenet interesting.
Gnutella is decentralized, but it's not anonymous, and therefore, it only does one of the things that make Freenet interesting.
|
|
|
|
|
|
|
|
| Latest Features |
|
| Latest CD Reviews |
|
|
|
